Previous Award Number: R305A100706
Previous Awardee: National Opinion Research Center (NORC)

Co-Principal Investigator: Jens Ludwig, University of Chicago

Purpose: The purpose of this study is to carry out a large-scale randomized efficacy trial of a case management program that focuses on reducing chronic absenteeism and improving school engagement by pairing mentor/monitors with students at risk for school dropout.

Project Activities: This study will conduct a truancy prevention experiment that randomizes at two levels – across and within schools – in order to identify the causal effects of the truancy reduction intervention on the target students and on the other students within the school. Chicago Public Schools students will be randomly selected to receive the program or to be in the control group. Within those schools selected to receive the intervention, eligible students will be randomly assigned either to receive the intervention or to be in the comparison condition.

Products: Products include peer-reviewed publications that present evidence on the main experimental impacts of the intervention on students. This will include an analysis of mediating mechanisms that lead to outcomes; evidence on any spillover effects; and a comprehensive benefit-cost analysis.

STRUCTURED ABSTRACT

Setting: This study will take place in Chicago Public Schools’ elementary and high schools.

Population: The intervention will be piloted during the 2010-11 school year with a target population that consists of 9th graders and elementary school students who missed a pre-identified range of school days during the previous (pre-study) year.

Intervention: Truancy officers will work with students and their parents to carry out an initial home visit and complete an assessment of the child and family. Following this, regular meetings, mostly with students but sometimes with parents, will take place throughout the school year along with periodic home visits. The data gathered from these activities will be combined into a monitoring data sheet that will include information on student attendance, behavior, and academic achievement.

Research Design and Methods: The design will be a randomized control trial that randomizes at two levels – across and within schools – in order to identify the causal effects of the intervention on the target children and the “spillover” effects on other students within the school. For the main mentoring effect, the study uses a multisite trial with a two-level design that places students within blocks. Each treatment school represents a block and within each school eligible students are randomly assigned to receive the intervention. For the “spillover” effect, outcomes of randomly-assigned control students in randomly-assigned treatment schools will be compared with outcomes of students in randomly-assigned control schools.

Control Condition: Students in the control schools and control students in the treatment schools will continue to receive the standard curriculum and other supports offered by the school district, and will not receive the intervention.

Key Measures: The Chicago Public Schools student-level administrative records will provide information on school attendance, grades, course credits completed, disciplinary actions, and scores on standardized achievement tests. To measure non-schooling outcomes, the researchers will request permission from relevant state and federal agencies to also examine individual-level administrative data on juvenile arrests, formal labor market involvement as measured by quarterly unemployment insurance records, and Medicaid records on eligibility and service utilization. In-person surveys will be done with a randomly selected sub-set of youth in the treatment and control groups consisting of: (a) very short surveys of education-related matters and mental/physical health for students including completion of the 33-item Student Engagement Instrument, plus reading and math assessments, and (b) short surveys of parents that focus on family routines and dynamics, mental and physical health, income, and receipt of social services.

Data Analytic Strategy: The causal effect of the intervention on the targeted truant students will be estimated by comparing the average outcomes of treatment versus control students within the grades and schools that are randomly selected for the intervention (the intention to treat, or ITT, effect). By randomizing at two levels (schools and students), researchers will also identify the “spillover” effects of increasing school attendance by previously truant students through analysis that compares the outcomes of students within treatment schools who do not themselves receive the intervention to students in randomly assigned control schools.

** This project was submitted to and funded under Education Policy, Finance, and Systems in FY 2010.